Graduating Microenterprises to SME-Level Credit

This project investigates some of the major obstacles to firm expansion from microenterprise to small and medium enterprise (SME) with a randomized experiment on Micro-Finance Institutions (MFIs) in South Africa. Like many LICs, South Africa’s prospects for large future injections of Foreign Direct Investment (FDI) are modest at best. As a result, the country must grow from within. And yet the country’s SME sector languishes, with start-up businesses unlikely to mature into new firms, and new firms unlikely to become established firms (GEM 2005). The research team hypothesizes that SME growth is constrained by a lack of credit caused by heterogeneity in the returns to investment combined with a market failure. While we might expect heterogeneity to be typically overcome as lenders “start small” and build a relationship with potential SME-level clients, this could be limited by a credit market friction common in many developing countries where micro entrepreneurs usually seek loans from subsidized MFIs which tend to adhere to a social agenda of helping the “poorest of the poor” rather than pursue a profit-maximizing strategy. These factors imply two things for the market equilibrium. First, MFIs will cross-subsidize low-return clients with profits from high-return clients and thus strive to hold on to their existing high-return clients. Second, unsubsidized lenders that wish to make SME-level loans will find it difficult to learn about micro-credit clients because they cannot compete with subsidized MFIs to pursue a “starting small” strategy.

The researchers will test this theory via a randomized experiment designed in partnership with an SME-level lender in South Africa. The lender will offer an incentive to loan officers from three MFIs to identify which of their clients are best suited for SME-level loans, and provide SME-level loans to a random selection of the clients listed. The incentive will be paid to the loan officer and client if they are selected to receive a loan, and if they successfully repay. This should first allow the team to answer two basic questions:

  • Do MFIs have private information on the transformational potential of their clients and, as a consequence, can the “starting small” approach work?
  • Can we reliably pinpoint firms that are likely to succeed at higher loan amounts?

Additionally, this study will also develop four main measurement methods that enable a better understanding of whether it is possible for informal microenterprises to be a dynamic source of growth:

  • The team will measure the returns to enterprises in the transition from micro to SME – as their cooperating lender is prepared to make loans averaging USD 5,000 (an SME-level loan size);
  • The experiment allows a simple method to estimate whether there are robust heterogeneous returns to investment, and to determine (ex-ante) the distribution of potentially “transformative” entrepreneurs in a population;
  • The team will examine whether entrepreneurs are biased in their beliefs, a possibility that would lead other experiments to systematically underestimate the potential for growth from microenterprise to SME;
  • Their measurement process will provide estimates of productivity enhancement that are independent of what is assumed about market structure, thereby deviating from most studies that focus on firm profitability.

This study’s primary outputs will include three waves of survey data collection, with 900 enterprise observations in each wave, which will be combined with consumer welfare data and administrative data from the cooperating lender. The experiment will focus on female entrepreneurs due to the fact that the cooperating lender believes the credit market friction at the heart of this study to have a greater effect on female entrepreneurs than male entrepreneurs.

The results of this study should be useful to both private and public sector entities in South Africa that are involved in the SME credit space, a handful of which the researchers have already established relations with. On the private sector side, these include the participating lenders to this study as well as the SME credit divisions of the large commercial banks. On the public sector side, the research team will target state-owned organizations such as the National Youth Development Agency, Khula Enterprise Finance, and the Small Enterprise Development Agency.


Dean Karlan

Northwestern University

Gharad Bryan

London School of Economics